Sabato's Crystal Ball

The Future Shape of the House

Kyle Kondik, Managing Editor, Sabato's Crystal Ball January 9th, 2014

The U.S. Census Bureau recently released new population projections for all 50 states. The results are not particularly surprising: American population growth is faster in the South and West, as it has been for decades.  In the U.S. House and the Electoral College, this growth makes the reallocation of seats every 10 years a zero-sum game. There are winners (generally states in the Southern and Western Sun Belt) and losers (usually in the Northern and Midwest Frost Belt).

Sean Trende of RealClearPolitics crunched the numbers, and his best projection is for Illinois, Michigan, New York, Ohio, Pennsylvania, Rhode Island and West Virginia to all lose one seat (and, therefore, one presidential electoral vote) apiece, while Colorado, Florida, Montana, North Carolina and Virginia would add one. Texas would add two.

From a presidential standpoint, these changes would slightly move the Electoral College in the Republicans’ direction: If this apportionment had been in place during the 2012 election, President Obama would have won 329 electoral votes, as opposed to the 332 he actually received. That continues a trend: Obama’s same 2012 result would have netted him six more electoral votes if it was conducted under the electoral map used in the 2004 and 2008 elections. This makes sense: the Northeast and much of the Midwest have been bluer in presidential elections than the South and Interior West.

Trende, as well as Daily Kos Elections (which also took a close look at the census projections), note that by the time 2020 rolls around, there might be other changes: California could gain a seat — as it did in every reapportionment from 1850 to 2010, when it failed to add a seat. Minnesota, which came very close to losing a seat to North Carolina in the most recent reapportionment, could end up a loser this time.

There are still years to go, of course, and unforeseen events can upset this process: 2005’s Hurricane Katrina helped cost Louisiana a House seat after 2010, for instance.

Reapportionment gets a little muddier when assessing what the changes in the number of seats each state gets would affect the House. Here, the swaps from the Northeast to the South and West might actually be beneficial to Democrats, even if they wouldn’t necessarily be in the Electoral College.

Take a look at Table 1, which shows information about the House delegations of the seven states that would lose House seats if these projections come true.

Table 1: The makeup of states that could lose seats after the 2020 census

These seven states feature four of the most pernicious gerrymanders in the country: Gov. Pat Quinn’s (D-IL) razor-thin victory in 2010 gave complete control of that state’s redistricting process to Democrats, which allowed them to lay the groundwork, through redistricting, to beat four Republican incumbents (three were freshmen elected in 2010). That was half of the eight seats Democrats netted nationally. Illinois also lost a seat because of slow population growth (as it is projected to again), and therefore Democrats eliminated another Republican seat.

Meanwhile, Republicans took total control of Michigan, Ohio and Pennsylvania in 2010 and netted a combined 12 House seats from these states (nearly a fifth of their 63-seat gain nationally that midterm). After the post-census redraw, not only did Democrats fail to recapture any of those seats, but they actually lost another in the general election (the seat now held by Rep. Keith Rothfus, a Western Pennsylvania Republican). Republican map-drawers also eliminated a Democratic seat in Michigan and Pennsylvania, while Ohio Republicans (guided by House Speaker John Boehner of the Buckeye State) swallowed hard and axed both a Democratic and a Republican seat in order to build a secure gerrymander for the rest of the GOP incumbents.

As Table 1 shows, Democrats are tightly packed in a small number of districts in these three states. Team Blue holds only 14 of 48 seats in these states, and those districts gave an average of 71% of their votes to Obama in 2012. Meanwhile, in the 34 Republican seats in those states, Mitt Romney only won 54% of the vote. Even though Democrats dilute themselves by being densely packed together in cities — versus the more suburban/rural Republican base — gerrymandering is still a big culprit in why there’s such a stark difference in these three states between their presidential preferences (all three voted twice for Obama) and their heavily Republican congressional delegations.

Obviously, we don’t know who will control redistricting in these three states come 2021. But even if Republicans retain total control, they may have a hard time finding Democratic seats to eliminate if they do in fact each lose a seat. There are only so many ways one party can slice and dice an electorate. Illinois Democrats, if they retain control of the process in their state, might face a similar problem.

One interesting wrinkle with these census projections is that some of the smallest states could lose seats. Rhode Island, which other than right before the first reapportionment (1790) has never had fewer than two seats, might dip down to one. West Virginia has never had fewer than three seats, but it might only have two in the next decade. These seat losses would probably cancel each other out. Rhode Island, with two heavily Democratic seats, would have to shed a Democrat. The Mountain State, meanwhile, has two Republicans and one Democrat, but given the state’s political trajectory — growing more Republican seemingly by the year, even down the ticket — it’s not a stretch to think that the state’s three House seats will all be Republican by then.

Table 2 shows the states that are currently projected to gain seats.

Table 2: The makeup of states that could add seats after the 2020 census

Notes: *Texas would add two seats, not just one, under this projection. ^Montana has only one seat, so it had no congressional redistricting after 2010.

The maps in these states are also mostly drawn by Republicans. For instance, the GOP drew maps in Florida and Virginia, which went for Obama twice, and North Carolina, which went for Obama once (in 2008). Partially as a result, Democrats only control 17 of 51 House seats (33%) in these states.

The growth in these states is disproportionately coming either from out-of-staters moving to job magnets like Northern Virginia, Raleigh-Durham and Charlotte or from minority growth: All three states have electorates that are more nonwhite than the national average.

Again, even if the GOP retains total control of redistricting in these states, it will be challenging for them to add a seat apiece and figure out a way to make it Republican, let alone protect the incumbents who are now benefiting from the current maps. The same factors that have been making these states more Democratic-leaning over time would also presumably make the job of redistricting even more difficult for Republicans.

Texas — slated to add two seats — is and will likely remain deeply Republican at the presidential level. However, it added four seats after 2010, and three of those four went to Democrats because of nonwhite population growth (and judicial intervention). So don’t just assume that two new seats in Texas means two new Republicans in Congress.

Finally, a new seat in Montana could be politically interesting. Montana — 55% for Romney in 2012, and slightly under 50% for John McCain in 2008 — isn’t as Republican as it seems. The state’s two biggest cities, Missoula in the west and Billings in the southern/central part of Big Sky Country, are political opposites. Missoula County cast about 57,000 votes for president in 2012, and gave 57% of those to Obama. Yellowstone County (Billings) cast almost 69,000 votes, and gave about 59% to Romney. Combined, those two counties accounted for about a quarter of all the votes cast in the state. If one of these districts was anchored in Missoula, it could be more competitive than the state as a whole.

Anyway, this is all hypothetical at the moment. So much can change, including the current makeup of the congressional delegations from these states, the census projections and which party controls redistricting (or whether the process becomes nonpartisan in a given state). But it’s interesting to think about the possibilities — especially from the perspective of the Democrats, who are currently frozen out of the House and might be so for the next few cycles, barring a major wave or, after 2020, a shuffling of redistricting and reapportionment.

More House members head for the exits

In our last Crystal Ball before the break for the holidays, we predicted there would be more House retirements in 2014 after a flurry in December. Only six days into the year, we were proven right: Rep. Jim Gerlach (R, PA-6), a moderate from the Philadelphia suburbs, called it quits. Then, yesterday, two Democrats joined him: Reps. Mike McIntyre (D, NC-7) and Carolyn McCarthy (D, NY-4).

McIntyre’s retirement moves his seat from Leans Democratic to Likely Republican: This is a seat where Romney took 59% of the vote in 2012. North Carolina Republicans made this district much more Republican in the post-2010 redraw, and while it did not allow them to defeat McIntyre in 2012, they’re very likely to take the seat now. Former state Sen. David Rouzer (R), who came within 0.2 percentage points of beating McIntyre in 2012, is running again here. The retirements of McIntyre and Rep. Jim Matheson (D, UT-4) open up two of the nine Democratic seats won by Romney in 2012, and both are likely to flip to the Republicans. So the real number of pickups the Democrats need to win the House is more like 19, as opposed to the more frequently reported 17.

Meanwhile, McCarthy’s Long Island seat (56% Obama) is could be competitive, so we’re moving it to Likely Democratic, with the potential for further movement toward Republicans depending on who emerges to run here and whether the national party views it as a top target.

Gerlach had a number of tough races during the last decade, but Republicans in Pennsylvania — as alluded to above — redrew the state’s maps to make his seat more Republican. Under the current lines, Obama got 53% in 2008, but only 48% in the more competitive 2012. Other prognosticators are calling this race Leans Republican, and they might be correct to do so given the changes made by Republicans to the district and the overall 2014 midterm political outlook (which looks decent for Republicans). But because we do not know which candidates will emerge, and because open seats are generally much more competitive than ones held by incumbents, we’re changing this race from Safe Republican to Toss-up. There have been several similar districts that have opened because of Republican retirements over the past few months: IA-3, NJ-3 and VA-10 are others. We’re calling all of them Toss-ups at the moment. Of seven Republican seats rated by the Crystal Ball as Toss-ups, five are open (FL-13, which is holding a special election in March, is the fifth).

Speaking of redistricting, if Democrats were to win and hold PA-6 for several cycles, and if Republicans got another chance at a post-2020 redraw, they could just give up on the district, turn it into a Democratic district, and shore up their other incumbents. If this thought hasn’t crossed the mind of Pennsylvania Republicans yet, it will.

We suspect we’ll have other House adjustments coming soon, but in the meantime, watch out for more retirements on both sides of the aisle. Now that the joy of the holidays is over, lawmakers returning to Washington must be asking themselves: Do I really want two more years of this?

Table 3: Crystal Ball House ratings changes